DIY RGBW Question

[Austin28]

Hello,

 

Can anyone tell me if these can work together? I am a little confused on how I would wire this and what sort of power supply I would need? (12V 1A or 24V 1A, or something with a slightly higher voltage). What's the deal with the positive output? That's would power the white correct? The remaing channels used for the rgb diodes, right. Please help?

 

http://www.amazon.com/Programmable-LED-Time-Controller-Multi-function/dp/B00NOQ9FP8

 

http://www.amazon.com/gp/aw/d/B00UOF7TKU/ref=mp_s_a_1_1?qid=1435299135&sr=8-1π=SL75_QL70&keywords=rgbw+led+chip

 

thanks!!

[evilc66]

Those two are not directly compatible. The controller is set up for common anode, meaning that all the positives for each color are tied together, and the controller switches the ground side (cathode) to change the output via pwm. You need an LED setup that is configured for this type of use, which the XM-L is not. It's to do with the voltage required of the LEDs. Common anode RGBW LEDs will typically have small resistors internal to the LED so that the average voltage for each channel is the same, which is required for a common anode setup. You could add resistors to the XM-L, but it's not really worth the effort. You would be better off using a true constant current driver (like Meanwell LDDs) for each channel and hook them up to an appropriate controller.

 

Out of curiosity, what are you planning on using the RGBW XM-L for anyways?

[Austin28]

That makes sense, thanks, still learning about LED'S. I was going to potentially use it for a reef tank (I am in the process of planning my first one) but I have also kept herps for a long time and have a collection of dart frogs and vivariums so I thought they would be neat for that too. A lot of my vivarium plants need LOTS of light to keep their best shape and color, but usually you can't find high powered rgbw light fixtures prebuilt and programmable for cheap, so I have gone with adding rgbww tape to existing 6500k light fixtures in the past. Thought this may have been an economical solution, lol.

 

In regards to the reef tank, I want to set up a fluval spec v and modify the stock light components while keeping the stock look of the fixture. Maybe I will do two channels only with a meanwell driver/controller, one white channel and some red, violet, warm white, deep red (or something along those lines) on the second channel. That seems pretty economical too. I just really wanted the full blown rgbw programmability for the relatively cheap price tag those components would have fetched.

 

Seems like you know a lot about electrical. I do not. I have only really begun working with it in my hobbies recently. Any good suggested reading out there? Something that touches on the more technical side of things but stars at a beginners level.

 

edit: Just read your other post, males me want to just do 1 channel of neutral white and on channel with ww and a little violet, blue, and cool white.

 

Any comments and questions are appreciated!!

 

Thanks

[evilc66]

RGBW sounds good on paper, as you can create almost any color of light easily. While that's great for the viewer (most of the time), it's not always the best for the organisms kept under them, especially for a reef. The peak wavelengths used in those LEDs are usually not the more desireable ones. For example, blue is usually 470nm (will make a reef tank look like windex), green is usually 525nm (makes the light brighter, but there are better options out there for that), and red is usually 630nm (adds warmth, but doesn't blend well with everything else).

 

For a planted/herp tank, a cool white (6500K), warm white (3000K), royal blue (455nm), violet (420-430nm), and lime (broad spectrum green) combination (not 1:1) is a good mix that will make everything pop from a color standpoint, and lots of growth potential.

 

For a reef, there are a few ways to do it based on how complex you want to make it. Ideally, you want to have as many controllable channels as possible so you can tweak the light to your hearts content.

 

A simple reef setup would be something like neutral white (4500K), royal blue (455nm), blue (470nm), and violet (430nm). Color rendering will be good, as will growth.

 

A more complex setup for a reef would be neutral white (4500K), warm white (3000K), royal blue (455nm), blue (470nm), cyan (490nm), lime (broad spectrum green), and violet (mix of 420nm and 430nm). That will give you the best coloration possible, and lots of growth.

 

The ratios for all these combinations will vary depending on your setup, and your desired look. While having individual control of each color is great, it does add cost and complexity. There are many of the colors that can be combined safely to reduce cost and complexity. A great premade DIY solution would be to use the v3 arrays from Nano Box Reef. Color is about perfect.

 

As for retro-fitting the Spec V light with something better, you are going to run into thermal issues with the stock fixture if you are going to upgrade to something more powerful.

[Austin28]

Any issues with an AI prime as another option?

[evilc66]

Not really. Seems to be a good fixture for the money.

[Austin28]

What would happen if I took the ped from the OP and I soldered all of the anodes together to a 1 amp 12v power supply? Could I make a common anode setup for one of these leds with the controller from the OP?

[evilc66]

If you soldered all the anodes together on that LED and wired it to a 12v power supply, you would end up with a very dead LED.

 

Really, you want to run LEDs in a constant current mode, not constant voltage. There are very few constant current drivers that can handle a common anode/cathode setup, and are generally run separately from each other (i.e. each color will have it's own driver and will not have any common wiring to any other channel).

 

Common anode setups are usually seen in cheap low current LED strips that use a constant voltage topology, and use PWM to alter the output of the LED. In RGB setups like this where the forward voltage on each channel is different, a resistor will be put in place to drop enough voltage on the relevant channels so that the effective forward voltage on each channel is the same (they have to be the same or the current draw on the lower higher voltage strings will be far higher than the others). While this works, it's not efficient, as you are burning off the excess voltage in heat. It's not a huge problem when you are talking low current LEDs, but with big high power LEDs like that XM-L, the amount of heat that you can burn off could be significant.

 

EDIT: Fixed a small mistake

[Austin28]

... Disregard this I didn't read the whole above post first time through. That's a very good explanation. I appreciate it!

[Austin28]

"they have to be the same or the current draw on the lower voltage strings will be far higher than the others"

 

Can you please explain this in more detail?

[evilc66]

Well, I did end up getting it backwards. The higher voltage string will pull more current. It's all to do with the fact that LEDs are current driven devices. If you look at the datasheets for common LEDs like the XP-E, XP-G, Rebel, etc... you will see that the relationship between voltage and current is not linear. Small changes in voltage can lead to large changes in current. In the example of an XP-E from Cree, at 3v, the LED requires about 225mA. Move the voltage up to 3.2v and the current soars to almost 600mA. This obviously leads to a brighter LED (that will be out of balance with the others in an RGB or RGBW setup), and a lot more heat. To compound the issue more, the LED's internal resistance changes as it heats up, so that current could increase again as the LED heats up. This really is why LEDs should be current driven. Lock the current, and the LED is stable regardless of the electrical conditions.

 

With an RGBW setup, having the same voltage across all the colors can create this imbalance. Each color has a slightly different forward voltage when driven at the same current. For the XM-L @ 350mA the white and blue dies run at 3.1v, the green die runs at 3.3v, and the red runs at 2.25v. If you were to run everything at an average of 3v, the white, blue, and green dies wouldn't have a problem, as they would be slightly underdriven (white and blue at ~215mA, green at ~150mA), but the red would die a very quick death. Resistors can be placed in series with each color to drop additional voltage (more so for the red than the others) so the effective forward voltage is closer to the other colors.

[Austin28]

Is there any way to calculate how the amperage is split in parallel wiring?

 

Also can you explain to me why the I-V curve for led's works the way it does? I get that it is exponential but what causes the amperage to be magnified so much as the voltage increases? I hope I phrased that question in a way that makes sense.

[evilc66]

The exponential V-I curve is just due to how the die is made, and is a characteristic of all LEDs regardless of size and power (it does change slightly from color to color and manufacturer to manufacturer, but the difference is small).

 

As for figuring out the current, you need to figure out what the voltage drop across each LED will be (source voltage divided by number of LEDs for a constant voltage setup), then line it up on the V-I curve in the datasheet.

 

Really though, I'd highly recommend not going down this path for high current LEDs. Constant current drivers are cheap enough and reliable enough to not even consider constant voltage. Constant voltage should only be considered if you need to drive a huge number of LEDs in series, and reliability is not critical. You will gain the ability to easily control the LEDs, and have far greater output stability if you go constant current.

[Austin28]

Okay I am over the common anode/parallel thing, lol. Good advice and information

 

What's the deal with this driver? Can it be set up with a controller? Did some research and I am assuming this is a big fat no, there is another model (EUC-144QxxxDT) with a dimming cable but I think it dims the whole driver...?

 

 

http://www.ebay.com/itm/Inventronics-160w-Led-Driver-Constant-Current-Type-Quad-Channels-Model-EUC-1-/141705995383?pt=LH_DefaultDomain_0&hash=item20fe55e477

 

Inventronics 160w Led Driver Constant Current Type Quad Channels Model : EUC - 1

 

How in the heck would I ground a heatsink using this type of setup?

 

Thanks!

 

Also, how in the heck do these work. The diodes have a different forward voltages but are common anode, how do they split the current and avoid overpower/underpower?

 

http://www.ebay.com/itm/10pcs-3W-RGB-LED-3-CHIP4Pin-full-color-Red-Green-Blue-3-watt-LED-Bead-Lamp-/261744113488?pt=LH_DefaultDomain_0&hash=item3cf12a3750

[evilc66]

Correct. The DT is dimmable via 0-10v or potentiometer, and dims the entire driver (individual channels cannot be controlled separately), and the ST is fixed output.

 

You don't need to ground the heatsink. With a constant current driver setup, it would actually be a bad idea if you had a short to the heatsink, depending on the driver. It's really dependent on whether the driver IC is sourcing or sinking, but you generally want to keep the input and output side of the driver separate.

 

That LED works the same way as the XM-L with the anodes tied together. You have to put a resistor in series with each die to bring the forward voltage up to the same level so the current consumption is the same.